Sharp logarithmic Kakeya maximal bounds

For each direction, take the largest average of a function’s absolute value over unit tubes of radius δ\delta pointing that way. How does the norm of these largest averages grow as the tubes narrow?

Why it matters and what counts as progress

Why it matters: The growth rate limits how tightly tubes pointing in different directions can overlap. It also controls the volume of thin neighborhoods of sets that contain a unit segment in every direction.

What counts as progress: An upper bound must hold for every function in the stated class. A matching example shows that the dependence on δ\delta cannot be improved. Weighted sums of tube indicators give a closely related estimate.

How the problem developed

  1. 16 Jul 2026 · Matching construction [5]

    Rafael J. Fernández-Delgado and Mikael de la Salle Constructed Kakeya sets with a neighborhood-volume upper bound of the conjectured logarithmic order.

  2. 25 Oct 2026 · Local-smoothing argument [4]

    OpenAI Developed analytic ideas that Lewko and ChatGPT later used for the sharp bound.

  3. 15 Nov 2026 · Sharp upper estimates [1][3]

    Mark Lewko and ChatGPT Reported the sharp maximal estimate and weighted tube bound. Palomar registered a Lean check of both upper estimates.

AI contribution

The sharp Kakeya maximal inequality in all dimensions

For n≥2n\ge 2 and 0<δ≤10<\delta\le1, Lewko obtains the LnL^n Kakeya maximal bound with factor (log⁡(2/δ))(n−1)/n(\log(2/\delta))^{(n-1)/n}. The weighted estimate bounds the integral of (∑jaj1Tj)q(\sum_j a_j 1_{T_j})^q by a dimension-dependent constant times log⁡(2/δ) δn−1∑jajq\log(2/\delta)\,\delta^{n-1}\sum_j a_j^q, where q=n/(n−1)q=n/(n-1), the weights are nonnegative, and the unit tubes have radius 2δ2\delta and δ\delta-separated directions.

AI involvementAI-centered
ImportanceNotable
Technical verificationFormal (Lean)
Human understandingFew
Reported

Sources

  1. Lewko’s result and account of the ChatGPT work
  2. Pinned Kakeya proof and attribution notes
  3. Palomar check of the two Kakeya upper bounds
  4. OpenAI’s local-smoothing argument
  5. Fernández-Delgado and de la Salle’s Kakeya construction
Publication history